When an AI system does something consequential — calls a tool, spends money, escalates to a person, refuses a request — you need a record that still means something six months later, when someone asks what happened and why. An application log does not survive that question. Anyone with file access can edit it, and nothing in the file says whether they did. agentledger writes each action as a validated, content-addressed record in a SHA-256 hash chain. Editing any stored record, in any field, breaks the chain, and verify_chain() says so. What you get - Order assigned by the ledger, never claimed by the caller - Every record carries a digest of its own content - Idempotent appends: no silent overwrites of history - Queries scoped to an explicit ledger cut, so "no matches" is never returned as "this never happened" - Safe concurrent writers, and a torn tail from a killed process is quarantined rather than silently truncating history - One dependency, Python 3.10+, 43 tests, no network Try it free before buying pip install phb-agentledger python demo.py The demo records three agent actions, proves the chain, then hand-edits the stored file and catches it. What this licence is agentledger is source-available. It is free for personal, educational and evaluation use, including evaluation inside a company. This licence covers commercial use: production systems, products and services offered to others, and internal business operations. It is perpetual for the version purchased, covers unlimited developers in the purchasing organisation, and needs no negotiated agreement. It gives you a defensible record of what your AI system did. It is not legal advice and does not by itself make any system compliant with any regulation. PHB Studio